COMPANY
Perplexity
Overview
Perplexity is an AI-search and personal-agent company best known for Perplexity.ai (conversational search), Perplexity Comet (free cross-platform AI browser), and Perplexity Personal Computer (Max-tier desktop-resident agent). In April 2026, Perplexity moved Personal Computer from private preview to general availability on macOS, anchoring the $200/month Max subscription tier around a 24/7 ambient desktop agent and positioning itself as the most direct consumer-product competitor to Anthropic’s Claude Cowork.
Timeline
- 2026-04-17-AI-Digest — Perplexity Personal Computer launches in general availability on macOS for all $200/month Max subscribers. The product turns a user-supplied Mac mini (M4 $599 model recommended) into a 24/7 ambient AI agent that lives in the background across every app, invoked by double-tapping the Command key and responding to text or voice. Persistent access to local files, Gmail, Slack, GitHub, and an expanding connector list; sees the active app and surfaces context-relevant quick actions. Explicitly gated to the Max tier (not the $20 Pro tier), using Personal Computer as the Max price anchor. Direct consumer-product challenger to Anthropic’s Claude Cowork (GA earlier this month). The requirement of a dedicated Mac mini kept plugged in and awake is Perplexity’s answer to the “my laptop sleeps and my agent dies” problem that Anthropic solves via Routines cloud execution — positioning Personal Computer as a prosumer/power-user product, not an enterprise-deployment surface. Perplexity Comet continues as the separate free AI-browser consumer surface across iOS, Android, Windows, and Mac.
- 2026-06-04-AI-Digest — Announces a hybrid local/cloud inference orchestrator added as a feature to the existing Perplexity Computer product (not a standalone product, not a rebrand) at Computex on June 2, with Intel + Nvidia RTX Spark support, shipping in July. Decides per-task what runs on-device versus in the cloud; targets both enterprise (Computer for Enterprise tier launched at Ask 2026) and consumer. Honest framing: this is one product feature plus one new chip family (Nvidia’s N1X) pointing in the same on-device-inference direction — cloud still owns frontier-capability workloads and the bulk of revenue. The right read is “hybrid routing graduates from research demo to product feature,” not “the next leg of inference economics.”
Key Developments
-
Personal Computer General Availability: The April 16 rollout to all Max subscribers is the first time Perplexity has gated a capability solely behind the $200 tier, transforming Max from a usage-uplift plan into a product-bundle plan.
-
Dedicated Hardware Requirement: The recommended $599 M4 Mac mini spec is a novel consumer-AI architecture — the agent runs on always-awake user-owned hardware, not on Perplexity’s cloud or on the user’s primary laptop. It’s a practical answer to the “agent needs persistent execution” problem but meaningfully restricts the addressable market.
-
Max-Tier Anchor Strategy: By restricting Personal Computer to Max, Perplexity is using the new surface to justify the high price point against both Claude Max ($200) and ChatGPT Pro ($200) — all three of which now need a flagship agent feature to hold their tier economics.
-
Cowork / Claude Code Routines Competitive Map: The three consumer desktop-agent products that matter at the $200 price tier as of April 16 are Claude Cowork (Anthropic), Personal Computer (Perplexity), and, to a lesser degree, ChatGPT Plus’s Codex/Atlas combination (OpenAI). Anthropic’s cloud-substrate approach and Perplexity’s user-hardware approach are the two architectural schools competing for mindshare in this segment.
-
Hybrid Local/Cloud Orchestrator Added to Personal Computer (June 2, 2026; ships July): A per-task router deciding what runs on-device vs in the cloud, with Intel and Nvidia RTX Spark as named hardware partners. The substantive piece is that this is a feature added to the existing Personal Computer product — not a rebrand or new product — and one of two same-week data points (with Nvidia’s N1X superchip) on the hybrid-routing direction. Read as feature-level graduation rather than tectonic-shift evidence.